Reflections of an AGU “First-Timer”
In the past few weeks, and during the conference, I collected a lot of opinions about the American Geophysical Union (AGU) Fall Meeting. Some said it was their most important week of the year, others said it was unlike any other conference they attend, a conference on steroids if you will, and I think there was a general consensus that the event was downright exhausting. I can’t comment with the second point, as this was the first conference that I’ve attended; however, I agree that this was easily the most important week to my research to date, and, yes, I am exhausted.
The week was exciting. I got to meet a lot of people interested in eScience and informatics. There was a broad coverage of research topics in the Earth and Space Science Informatics sessions, covering nearly every aspect of data management and curation. There were presentations on the applications of data transfer formats, including netCDF and HDF, and conformance to metadata standards, such as ISO 19115. Beyond simple data and metadata representations, many were interested in interoperability techniques, including OPeNDAP for application-level interoperability of data, and also vocabulary-level interoperability, which I found was a much more difficult concept for data curators to find value in. Data curators are not yet seeing the importance of vocabulary interoperability because the “killer app” that utilizes integrated vocabularies is lagging far behind the initiatives towards interoperability. I find this to be surprising as the “killer app” that many are looking for is a simple portal for spatio-temporal data search. The major hindrances to the development of a spatio-temporal data search portal is not only metadata interoperability (through standards and vocabularies), but also search efficiency. Technologies like Apache Solr are emerging for distributed indexing and query for keyword search over the metadata of deep web holdings, but spatio-temporal indices are being included in these technologies as an after thought.
Beyond the data/metadata/vocabulary research presented at the conference, there were numerous presentations on the importance of standardized web services, and many applications demonstrating the use of standards developed by the Open Geospatial Consortium, Amazon, and Open Archives Initiative, to name a few. In the many presentations, I don’t think there was a clear indication of what the most popular web service standard was, nor was there an indication that the diverse set of available standards would converge. Actually, Dan Pilone and his collaborators from NASA’s Earth Observing System (EOS) Clearinghouse (ECHO) put together a nice set of posters comparing the utility of web service standards from the perspective of both the developer and the end-user. All of this work was particularly interesting to my research, which is starting to converge towards application- (or service-)level semantics.
My poster presentation was more inline with those focused on the applications and integration of web services. I presented S2S, an application level framework for building customized user dashboards for data search and visualization that supports techniques such as advanced search and faceted browse. The poster was well received, and I gathered both positive feedback and, to an extent, skepticism. I spent the afternoon explaining the benefits to be reaped from application-level semantics, the successful application of the Semantic Web Methodology and Technology Development Process, and, for the skeptics, justifying the complexities of an application ontology. I was able to compile the input of the conversations surrounding the poster, realizing the need to incorporate a data abstraction into the application ontology, and the need to provide further abstraction on the service concept to support a broader range of web services. I think such abstractions will help to dissolve interest in web service standard convergence, and could serve as a foundation to a “killer app” that will spark further interest in vocabulary interoperability, and application-level semantics.
The week was exhausting. I find that I need to work on my ability to absorb information, as even just days after the conference, while putting together this blog, I feel fuzzy on the details of the many conversations I partook in. One of the problems I faced was that I am not great at taking notes and participating in a discussion concurrently. In addition to this, I found it hard to get the chance to sit down and reflect on conversations had, both because some conversations would span an hour or more, and also because I found I was moving directly from one conversation to the next (too many things to do in too little time!). I plan to work on my ability to jot down notes in the midst of discussions, and will be weary of transitioning to new conversations without taking a break (to reflect) in future conferences. If any of these reflections on my experiences have taken your interest, feel free to shoot me an email (or find me in the lab, TWCers!).